This discussion (and the bykim draft) reminds me of some materials I briefed a long time ago during an NGTRANS session at IETF50:
http://6bone.net/ngtrans/IETF-50-Minneapolis/Templin-v6v4compat.ppt
(See slides #15-#19; note that the colors in the diagrams denote distinct IPv6 prefix assignments).
The breifing presented a high-level vision of how a hierarchy of IPv6 prefixes could be driven deeply into an enterprise network ([RFC1918], section 1) during the early phases of IPv6 deployment w/o delving into technical details save for one - ISATAP was then and remains today the basic IPv6-over-(foo) mechanism required to support the vision.
In terms of other required functions, this thread is touching on many but it seems surprising that folks are viewing things from an "either/or" perspective. Instead, it seems clear that the solution will not amount to picking one of DHCPv6 vs. IPv6 ND, but rather combining the best aspects of both.
For example, the notion of "stateless prefix delegation" seems a clear oxymoron; prefix delegations require servers to keep lease lifetimes and employ mechanisms to avoid duplicate assignments and reclaim expired prefixes. We already have the state engine of DHCPv6 for this, and so how could it make sense to re-invent this wheel vis-a-vis extensions to the stateless mechanisms specified by RFCs 2461/2? By the same token, router lifetime advertisements, autoconfiguration prefix advertisements, etc. are clearly stateless mechanisms and so how could it make sense to use DHCPv6 for this instead of IPv6 ND?
It seems like folks might be hung up somewhat on mistaking the actual format of messages sent over the wire with the protocols implemented at the endpoints. In other words, it just doesn't matter what we call the messages that are sent over the wire as long as they are kept to a minimum, are efficiently coded, and support the hybrid functions required for combining the best aspects of DHCPv6 and IPv6 ND. Perhaps this entails piggybacking DHCPv6 and IPv6 ND messages? Perhaps this entails deploying a DHCPv6 server/relay on each IPv6 router/ND proxy?
To use an analogy, the legendary golfer Bobby Jones said that the secret to shooting low scores is to "find a way to turn 3 shots into 2". If we can find a way to combine the best aspects of DHCPv6 and IPv6 ND, and "turn 4 messages into 2", then everyone wins.
Fred [EMAIL PROTECTED]
-------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
